Robust Sensitivity-Aware Chance-Constrained MPC for Efficient Handling of Multiple Uncertainty Sources
James Zhu, Thierry Simeon, Marco Cognetti
AI summary
Problem
Real-world robots must navigate environments with simultaneous uncertainties from state estimation, model parameters, and dynamic obstacles, but existing robust control methods are either overly conservative, computationally prohibitive, or unable to integrate diverse uncertainty sources efficiently.
Approach
The method propagates Gaussian uncertainty estimates from Kalman filters using sensitivity analysis and embeds them into a chance-constrained optimization framework, allowing the planner to dynamically adjust safety margins based on quantified future uncertainty.
Key results
- Propagates Gaussian uncertainty from Kalman-filter estimates via sensitivity analysis
- Integrates multi-source uncertainty into a tractable chance-constrained MPC framework
- Achieves significantly higher constraint satisfaction and robustness than deterministic MPC
- Maintains real-time computational efficiency through warm-starting and analytical derivatives
Why it matters
Provides a computationally efficient pathway for deploying reliable, uncertainty-aware robotic navigation on resource-constrained platforms in dynamic real-world settings.
Abstract
Robust motion planning under uncertainty is critical for unlocking real-world robotics applications. This paper intro- duces SupeR-MPC, a computationally-efficient, sensitivity-aware, chance-constrained optimization framework that systematically accounts for multiple sources of uncertainty, including state esti- mation error, model parameter uncertainty, obstacle localization error, and process noise. This approach advances sensitivity- aware robust control by integrating chance-constrained opti- mization to handle the uncertainty models of Kalman-filtering methods. To demonstrate robustness against multiple uncertainty sources, SupeR-MPC was validated on a range of systems and environments, from a simple 2D example to a multi-agent dy- namic obstacle avoidance scenario. Comparisons against existing MPC methods show that SupeR-MPC significantly improves constraint satisfaction and robustness while maintaining real-time computational efficiency. These results highlight the effectiveness of sensitivity-aware chance constraints in enhancing real-world robotic decision-making under uncertainty.